After reading the article, I do think he has a point to some degree. I've seen comments on here that refer to Unix as like the wheel - Linux is simply polishing an already near-perfect idea. This can't be further from the truth. Look at what the article said -- he was disappointed that Stallman chose to pursue an open source version of an OS that was already at the time recognized as being not only old, but flawed. Age isn't the issue - it's idiotic to toss out an idea simply because of age. But flaws and technical reasons are definitely a cause to reconsider things, such as Unix. It would have been nice if he had chosen to open up a more revolutionary idea and push the field ahead instead of stalling it somewhere in the early 1970s. This isn't a unique sentiment. Rob Pike said a similar thing a few years ago at a talk in Utah (http://herpolhode.com/rob/utah2000.pdf) where he lamented that the fixation on Linux/Unix was leading many people to have blinders restricting them to thinking about the world in a very limited, closed-minded way. Rob happens to come from Bell labs where the original Unix creators realized that more was possible than what the Unix model provided, and they created Plan9 and Inferno. The open source model, while good for code freedom, seems to breed more than anything an irrational devotion to specific technologies simply because implementations of them exist for free. Why are people unwilling to consider that there could be better ways to do something, and that Unix/Linux is not the pinnacle of perfection in operating systems and software? I use Linux every day and enjoy it, but wouldn't blink an eye if something better came along and Linux got tossed off my machine.
I think your critique is woefully out of date. You are correct if you limit the neural network to the basic neural network models of decades past. From what I've seen at conferences in the HPC world lately, the more recent models do more than just use capacity to increase the size and connectivity of the network, but take into account more realistic physical models such as the electrical properties of the brain and mechanisms by which signals propagate both within neurons and across synapses. You're not looking at just a bigger back propagation network with sigmoid nonlinearities here -- the neural modeling world has moved far beyond that, in part due to increased interest and participation of neuroscientists. Unfortunately, most CS folks fail to learn much about the current state of the art beyond the basics such as the material from Simon Haykin's text (which, mind you, is pretty good).
Or the other favorite, the single moron who doesn't know how to unsubscribe and sends "unsubscribe" to the list, followed by others who do the same, followed by people sending instructions to unsubscribe, followed by more "unsubscribe" messages by those who can't follow instructions.
For a beginning setup, I would recommend a simple Meade telescope such as the ETX series (I have one of these). The telescopes are motor driven so they can accurately target and track objects, and have remote controls that let you program in coordinates to go seek out -- very useful when coupled with either a book of easy to find targets (like those star guides most bookstores carry) or software that lets you find interesting objects to look at. Personally I prefer, when looking around for pleasure, to use software on my computer to get a sense for what is up at any given time in the night, and then work with a book and a red flashlight (gotta have one of these - red flashlight won't hurt your night vision like a normal flashlight, or for that matter, the glow from a laptop screen). Software, although useful, takes some of the fun and challenge out of learning how to find things in the sky on your own. Besides, part of the fun of the hobby is getting to see everything in the sky that you look at while you seek out your target - using the computer to find things means you miss out on all those things in between the named objects.
You actually can get into astrophotography for a relatively low price. The only big requirements are that you get a camera that supports long exposure times and has a mount for a detachable lens (or, can be mounted onto the telescope above an existing eyepiece). You can buy eyepieces for the telescope that can attach to the camera to do photography through the telescope. Now, you won't be making pictures that are very high quality of things like nebulae and galaxies, but you would likely be able to practice and get used to the process on easier, bright and big targets like the moon. Taking a good, crisp picture of craters on the moon is no simple task, and that alone is easier than, say, getting a good picture of the Orion nebula. Once you choose a telescope, you should do some research on google to find a mount that fits the camera you have. You can likely come up with something to make a $300 digital camera work with your telescope if you do a little research. With $1000, a simple setup at home can easily be made to let you learn about the sky and take a few pictures to top it off. You won't be making anything that will be making it into Sky & Telescope, but you can set something up to learn enough about the process so that you can decide if a further investment into a better telescope or camera will be something you'll be willing to commit the time and effort to learning how to use. This is a hobby that takes time and effort, so it's wise to limit yourself to something simple and digestable early on so you can guage if you will have the patience to push it further and make a larger investment in a better setup.
"Also, this may be a stupid question, but I wonder how one measures the 'randomness' of a generator?"
There are lots of places that talk about this. A simplistic explanation of what it means to be a good PRNG is simply to provide a sequence of numbers with no correlations that matches the desired distribution. (http://mathworld.wolfram.com/RandomNumber.html). Books on modeling and simulation often have good explanations of this. This page (http://csrc.nist.gov/rng/) has a good overview, including simple descriptions of 16 statistical tests of interest.
It's sad to say, but that is a horribly naive point of view. Open source simply means that the source is available. It does not come with a stamp of approval that it has been somehow certified as being 'safe' or 'not evil'. Sure, a small project might be read over by a hobbyist out of curiousity, but for a large project with tens or hundreds of thousands of lines of code, I personally do not at all believe that some benevolent third party has sifted through both a massive code base and the corresponding complex control, data, and logical flow of the code. I would assume that any sufficiently evil chunk of code would be both uncommented and somewhat obfuscated to hide the true intent of the code. So, it would be like finding a needle in a haystack, only after the needle has been painted hay colored. Open source does not imply safe - it simply implies open source code. You are correct that it has the POTENTIAL to be safer than closed source, but there is absolutely no guarantee that it has been certified as safe/not-evil/whatever.
I'm not a big fan of eclipse (although I've tried it over, and over, and over, hoping one day it would make me "see the light" with some new, nice version). I wouldn't say it's popularity is due to a license though. Eclipse has momentum due to it being hyped as being representative of "industry best practices and standards". Some people are quite happy to use that as a metric for deciding to adopt it, regardless of whether those "best practices" apply to their particular programming problem or industry, or if the tool is actually technically superior to others available for the problem domain. This is a pervasive attitude in computing - there is an unusual and hard to explain aversion to having a diversity of ways to solve a problem. It's very much a "all you have is a hammer, so everything looks like a nail" mentality to me. I say use the right tool for the job, be it a language, framework, programming pattern, or editor. Choosing a "standard" because it's all you know or are comfortable with, and contorting it to do something it wasn't designed for is one of the contributing factors to all of the bad software that exists in the world.
This reminds me of a clever optical sorting algorithm I ran across a paper on in recent years (see http://www.cs.auckland.ac.nz/CDMTCS//researchrepor ts/244dominik.html). Again, a clever thought experiment - not sure how feasible it will be anytime soon to actually use though.
Is it just me, or is messaging in a social network not that much different than e-mail with an address book that the social network site manages for you? We're not talking a huge difference from e-mail here (other than less unsolicited spam). I'm not sure this means e-mail is dead - more that it's evolving. Quite honestly, if SMTP-based mail dies and an alternative with less annoyances (spam) pops up, great. But it's still just e-mail with a different protocol underneath...
I'm surprised I haven't seen anyone else mention this point. I have a little personal server here at home, but still plan to use remote server services like google or.mac simply because it makes me very uncomfortable to have all of my important data in one physical location. A fire, a flood, a thief - whatever the cause, having everything in one place makes me nervous. So, I like having an option for storing stuff that isn't colocated with me to cut the odds of important data being lost. This will always be a selling point for a remotely hosted server, regardless of how capable home solutions get.
Call me crazy, but that just might fall under the "negative impact on the environment" category. I'm not sure replacing CO2 producers with a big hole and a cloud of really nasty fallout would really do the trick.
And what moron modded the parent to this "insightful"? Hopefully that one gets fixed in metamoderation...
I've complained plenty to Subaru about the lack of options for my Outback. Up until the '07 models, the only solution was the less-than-optimal FM transmitter or an aftermarket replacement for the factory stereo. Unfortunately, in the '07 model they gave it zero effort in "integration" - all we get is an AUX input plug. Sure, this has the benefit of not being iPod specific, but it requires that I have an additional piece of hardware between the iPod and the car to take unamplified (ie, not from the headphone jack) output from the iPod and to feed into the car. Otherwise I'll have the fun that I used to have with the cassette deck adapters where you have to find that nice balance of cd player/iPod volume and system volume. Subaru is near the bottom of the heap with respect to iPod or Satellite radio integration, relying on external units with FM or (now) a single AUX plug as connection. There are many other manufacturers who have done it much better than Subaru has in terms of getting the audio signal into the stereo, and integrating control of the radio/iPod into the car controls to reduce safety risks. For a daily long distance commuter, this is definitely impacting my next car purchase. It's really a shame that such a nice car on the driving end of things fails so miserable in features like this.
"Benchmarks of any multiuse system are never universal. They best they can do for a large list like that is to use a benchmark that can reasonably represent a common use of such systems."
The linpack benchmark used to do the top500 list is a basic, dense matvec solver algorithm. (See wikipedia : http://en.wikipedia.org/wiki/LINPACK) This algorithm used to be the core of most scientific codes, back in the days when you would simply use the computer to solve a simple (but large) set of equations. In the last decade(s), the scientific world has moved to unstructured problems where the solvers are no longer solely matvec operations. Adaptive mesh methods, multigrid, and other similar "modern" methods in scientific computing do NOT have the same behavior as a basic dense matvec - a simple case would be considering a matvec problem where one deals with sparse matrices. Life gets even worse if you try to use linpack to reason about how a machine would perform on something highly data dependent, such as an n-body code or molecular dynamics simulation.
Linpack is really an archaic relic of the past, and it is NOT a benchmark of a multiuse system. It is a benchmark of a supercomputer from 15+ years ago. This is not news in the parallel computing world -- many efforts such as ParkBench, NAS Parallel Benchmarks, Livermore loops, etc... have been proposed as replacements for linpack to better cover the sorts of applications that a real "multiuse" systems will run. Unfortunately, the fact that most procurement folks and politicians who help fund these big govt. machines do not understand that linpack is a total waste of time have caused it to persist, contrary to the desires of people who either use the systems, or spend their careers studying performance issues in big parallel systems.
The fake elements originated at Lawrence Berkeley Lab, NOT LANL. (See http://www.sciencenews.org/articles/20020720/fob5r ef.asp) And the mustang story is largely false, although the mainstream press did not make a big deal out of the fact that the story originally reported turned out to be untrue. (See http://www.lanl.gov/news/index.php?fuseaction=home .story&story_id=1453). Also keep in mind that that LBL, Argonne, Brookhaven, etc... do minimal amounts of classified work compared to LLNL, LANL, and SNL. Even PNL and ORNL do significantly less than the big three. So if LANL, LLNL, and SNL tend to have more security incidents, then one cannot ignore that a stellar record from a laboratory that does no classified work means very little in comparison.
Please get your facts right. It's that sort of uninformed, incorrect rhetoric and accusation that got LANL in the press-generated hot water it currently finds itself in. Are you a politician?
Odd -- I've seen the opposite when I've meta-moderated in the last few months.
That, and people who appear to have very odd senses of humor judging by some of the +1 Funny mods out there...
I like iTunes also. It has a few features missing that I'd like, but I doubt they are interested in adding things that cater to folks with 100+GB music libraries.
iTunes is good.
This sort of thing is the typical FUD that emanates from OSS zealots. An iPod, using music ripped from CDs to either MPEG-4 audio (AAC, or Advanced Audio Coding) or Apple lossless, gives both high quality AND freedom. If you own an iPod, it's not likely that you care about interoperability with other players (unless you give your music away). If you rip your own CDs, you don't have "DRM-laden music". I dislike iTunes as much as the next person, although not for DRM reasons (I like the physical CDs as "backup").
See, iPod != iTunes music store. It's a perfectly reasonable audio player that on it's own does NOT impose DRM rules on you (other than not being able to easily copy files OFF on someone elses computer). So it lacks OGG support -- yes, OGG sounds good (I don't know what the previous poster was smoking - OGG sounds fine). The iPod doesn't force you into anything even related to the iTunes music store if you don't want to use it.
And why would anyone want a "wealth of codecs"? I can't see why someone wouldn't just choose one, and stick with it, and possibly use a higher quality second choice for a subset of their music. My iPod supports a "wealth" of codecs already anyways - just not OGG.
"The only difference being that one group is employed by Microsoft, and the other likely do it in their spare time (they're usually still employed by some big-ass technology company)."
Many of the big OSS projects (Linux kernel, Eclipse, as two huge examples) are predominantly written by people paid to work on the projects. The only difference between those hackers and one at microsoft is the openness of the code - not spare time vs work time. I think the "OSS = programmed in my spare time in the evening" argument went out the window a few years back for most of the truly large projects out there.
Notice that the google count drops when you remove references to web pages that, although you are filtering to only show results in English, still show pages that reflect the existence of the word "boxen" as meaning "boxes" in German. That, and the various people with the last name Boxen, and reverences to books with the word Boxen in them related to C.S. Lewis, and so on.
Using google hit counts is about as dumb as thinking that boxen is a real word simply by sticking a B on the front of the plural of Ox. "Common usage"? You do realize that the majority of computer users out there still refer to their computer as a "computer", not a "box" - let alone worrying about the plural of a somewhat niche-oriented, esoteric bit of terminology. To use your ill-informed google-metric to try to see how the "general population" uses the "word" "boxen", pages with references to boxen in the context of windows number in the mere 20,000 count. And I would speculate that many of those are uses by linux nerds who are in some way making nasty comments about Windows (disclaimer: linux user here. no windows junk in my space.)
Shall we remind you of the word "Jargon"? Existence in the world of jargon does not imply accepted use in the larger language.
Besides, FPGAs have two issues that make them good only for a very specific set of apps. Number 1, they don't currently have great floating point performance - this is a killer for most scientific apps. Number 2, they are hard to feed because the rate they can compute at versus the rate memory can feed them is quite skewed. Regardless, they're still very promising. The reconfigurable computing team at LANL (http://rcc.lanl.gov/) has done some very cool things with FPGA based systems.
Is anyone else frustrated when you hear wonderful science like this being done, yet see that probes like this are slated to have their funding cut (http://www.spacedaily.com/news/voyager1-05a.html) ? For some reason, $4.2 million / year to operate them (ie, listen) seems unbelievably cheap for such a unique resource - not only are there only TWO probes out there (voyager 1 and 2), but to get others out to replace them would cost a whole ton more....In addition to having to wait another 20 or so years to get there.
Science just doesn't work when politics gets involved...:(
Slashdot Mirror
After reading the article, I do think he has a point to some degree. I've seen comments on here that refer to Unix as like the wheel - Linux is simply polishing an already near-perfect idea. This can't be further from the truth. Look at what the article said -- he was disappointed that Stallman chose to pursue an open source version of an OS that was already at the time recognized as being not only old, but flawed. Age isn't the issue - it's idiotic to toss out an idea simply because of age. But flaws and technical reasons are definitely a cause to reconsider things, such as Unix. It would have been nice if he had chosen to open up a more revolutionary idea and push the field ahead instead of stalling it somewhere in the early 1970s. This isn't a unique sentiment. Rob Pike said a similar thing a few years ago at a talk in Utah (http://herpolhode.com/rob/utah2000.pdf) where he lamented that the fixation on Linux/Unix was leading many people to have blinders restricting them to thinking about the world in a very limited, closed-minded way. Rob happens to come from Bell labs where the original Unix creators realized that more was possible than what the Unix model provided, and they created Plan9 and Inferno. The open source model, while good for code freedom, seems to breed more than anything an irrational devotion to specific technologies simply because implementations of them exist for free. Why are people unwilling to consider that there could be better ways to do something, and that Unix/Linux is not the pinnacle of perfection in operating systems and software? I use Linux every day and enjoy it, but wouldn't blink an eye if something better came along and Linux got tossed off my machine.
I think your critique is woefully out of date. You are correct if you limit the neural network to the basic neural network models of decades past. From what I've seen at conferences in the HPC world lately, the more recent models do more than just use capacity to increase the size and connectivity of the network, but take into account more realistic physical models such as the electrical properties of the brain and mechanisms by which signals propagate both within neurons and across synapses. You're not looking at just a bigger back propagation network with sigmoid nonlinearities here -- the neural modeling world has moved far beyond that, in part due to increased interest and participation of neuroscientists. Unfortunately, most CS folks fail to learn much about the current state of the art beyond the basics such as the material from Simon Haykin's text (which, mind you, is pretty good).
Or the other favorite, the single moron who doesn't know how to unsubscribe and sends "unsubscribe" to the list, followed by others who do the same, followed by people sending instructions to unsubscribe, followed by more "unsubscribe" messages by those who can't follow instructions.
Say "CODEC" to the majority of consumers of this sort of technology, and get your answer in the blank stare that comes back.
For a beginning setup, I would recommend a simple Meade telescope such as the ETX series (I have one of these). The telescopes are motor driven so they can accurately target and track objects, and have remote controls that let you program in coordinates to go seek out -- very useful when coupled with either a book of easy to find targets (like those star guides most bookstores carry) or software that lets you find interesting objects to look at. Personally I prefer, when looking around for pleasure, to use software on my computer to get a sense for what is up at any given time in the night, and then work with a book and a red flashlight (gotta have one of these - red flashlight won't hurt your night vision like a normal flashlight, or for that matter, the glow from a laptop screen). Software, although useful, takes some of the fun and challenge out of learning how to find things in the sky on your own. Besides, part of the fun of the hobby is getting to see everything in the sky that you look at while you seek out your target - using the computer to find things means you miss out on all those things in between the named objects.
You actually can get into astrophotography for a relatively low price. The only big requirements are that you get a camera that supports long exposure times and has a mount for a detachable lens (or, can be mounted onto the telescope above an existing eyepiece). You can buy eyepieces for the telescope that can attach to the camera to do photography through the telescope. Now, you won't be making pictures that are very high quality of things like nebulae and galaxies, but you would likely be able to practice and get used to the process on easier, bright and big targets like the moon. Taking a good, crisp picture of craters on the moon is no simple task, and that alone is easier than, say, getting a good picture of the Orion nebula. Once you choose a telescope, you should do some research on google to find a mount that fits the camera you have. You can likely come up with something to make a $300 digital camera work with your telescope if you do a little research. With $1000, a simple setup at home can easily be made to let you learn about the sky and take a few pictures to top it off. You won't be making anything that will be making it into Sky & Telescope, but you can set something up to learn enough about the process so that you can decide if a further investment into a better telescope or camera will be something you'll be willing to commit the time and effort to learning how to use. This is a hobby that takes time and effort, so it's wise to limit yourself to something simple and digestable early on so you can guage if you will have the patience to push it further and make a larger investment in a better setup.
"Also, this may be a stupid question, but I wonder how one measures the 'randomness' of a generator?" There are lots of places that talk about this. A simplistic explanation of what it means to be a good PRNG is simply to provide a sequence of numbers with no correlations that matches the desired distribution. (http://mathworld.wolfram.com/RandomNumber.html). Books on modeling and simulation often have good explanations of this. This page (http://csrc.nist.gov/rng/) has a good overview, including simple descriptions of 16 statistical tests of interest.
It's sad to say, but that is a horribly naive point of view. Open source simply means that the source is available. It does not come with a stamp of approval that it has been somehow certified as being 'safe' or 'not evil'. Sure, a small project might be read over by a hobbyist out of curiousity, but for a large project with tens or hundreds of thousands of lines of code, I personally do not at all believe that some benevolent third party has sifted through both a massive code base and the corresponding complex control, data, and logical flow of the code. I would assume that any sufficiently evil chunk of code would be both uncommented and somewhat obfuscated to hide the true intent of the code. So, it would be like finding a needle in a haystack, only after the needle has been painted hay colored. Open source does not imply safe - it simply implies open source code. You are correct that it has the POTENTIAL to be safer than closed source, but there is absolutely no guarantee that it has been certified as safe/not-evil/whatever.
I dunno about that. It's not really authentic Spanish Slashdot until their poll has an option involving "VaqueroNeal".
I'm not a big fan of eclipse (although I've tried it over, and over, and over, hoping one day it would make me "see the light" with some new, nice version). I wouldn't say it's popularity is due to a license though. Eclipse has momentum due to it being hyped as being representative of "industry best practices and standards". Some people are quite happy to use that as a metric for deciding to adopt it, regardless of whether those "best practices" apply to their particular programming problem or industry, or if the tool is actually technically superior to others available for the problem domain. This is a pervasive attitude in computing - there is an unusual and hard to explain aversion to having a diversity of ways to solve a problem. It's very much a "all you have is a hammer, so everything looks like a nail" mentality to me. I say use the right tool for the job, be it a language, framework, programming pattern, or editor. Choosing a "standard" because it's all you know or are comfortable with, and contorting it to do something it wasn't designed for is one of the contributing factors to all of the bad software that exists in the world.
This reminds me of a clever optical sorting algorithm I ran across a paper on in recent years (see http://www.cs.auckland.ac.nz/CDMTCS//researchrepor ts/244dominik.html). Again, a clever thought experiment - not sure how feasible it will be anytime soon to actually use though.
Is it just me, or is messaging in a social network not that much different than e-mail with an address book that the social network site manages for you? We're not talking a huge difference from e-mail here (other than less unsolicited spam). I'm not sure this means e-mail is dead - more that it's evolving. Quite honestly, if SMTP-based mail dies and an alternative with less annoyances (spam) pops up, great. But it's still just e-mail with a different protocol underneath...
A square 185 kilometers on each side is nothing - that's approx. 0.02% of the surface area of the pacific. The ocean is big.
I'm surprised I haven't seen anyone else mention this point. I have a little personal server here at home, but still plan to use remote server services like google or .mac simply because it makes me very uncomfortable to have all of my important data in one physical location. A fire, a flood, a thief - whatever the cause, having everything in one place makes me nervous. So, I like having an option for storing stuff that isn't colocated with me to cut the odds of important data being lost. This will always be a selling point for a remotely hosted server, regardless of how capable home solutions get.
And what moron modded the parent to this "insightful"? Hopefully that one gets fixed in metamoderation...
...and no cell, which was the point of the parent poster buying the PS3 to run linux in the first place.
I've complained plenty to Subaru about the lack of options for my Outback. Up until the '07 models, the only solution was the less-than-optimal FM transmitter or an aftermarket replacement for the factory stereo. Unfortunately, in the '07 model they gave it zero effort in "integration" - all we get is an AUX input plug. Sure, this has the benefit of not being iPod specific, but it requires that I have an additional piece of hardware between the iPod and the car to take unamplified (ie, not from the headphone jack) output from the iPod and to feed into the car. Otherwise I'll have the fun that I used to have with the cassette deck adapters where you have to find that nice balance of cd player/iPod volume and system volume. Subaru is near the bottom of the heap with respect to iPod or Satellite radio integration, relying on external units with FM or (now) a single AUX plug as connection. There are many other manufacturers who have done it much better than Subaru has in terms of getting the audio signal into the stereo, and integrating control of the radio/iPod into the car controls to reduce safety risks. For a daily long distance commuter, this is definitely impacting my next car purchase. It's really a shame that such a nice car on the driving end of things fails so miserable in features like this.
"Benchmarks of any multiuse system are never universal. They best they can do for a large list like that is to use a benchmark that can reasonably represent a common use of such systems."
The linpack benchmark used to do the top500 list is a basic, dense matvec solver algorithm. (See wikipedia : http://en.wikipedia.org/wiki/LINPACK) This algorithm used to be the core of most scientific codes, back in the days when you would simply use the computer to solve a simple (but large) set of equations. In the last decade(s), the scientific world has moved to unstructured problems where the solvers are no longer solely matvec operations. Adaptive mesh methods, multigrid, and other similar "modern" methods in scientific computing do NOT have the same behavior as a basic dense matvec - a simple case would be considering a matvec problem where one deals with sparse matrices. Life gets even worse if you try to use linpack to reason about how a machine would perform on something highly data dependent, such as an n-body code or molecular dynamics simulation.
Linpack is really an archaic relic of the past, and it is NOT a benchmark of a multiuse system. It is a benchmark of a supercomputer from 15+ years ago. This is not news in the parallel computing world -- many efforts such as ParkBench, NAS Parallel Benchmarks, Livermore loops, etc... have been proposed as replacements for linpack to better cover the sorts of applications that a real "multiuse" systems will run. Unfortunately, the fact that most procurement folks and politicians who help fund these big govt. machines do not understand that linpack is a total waste of time have caused it to persist, contrary to the desires of people who either use the systems, or spend their careers studying performance issues in big parallel systems.
The fake elements originated at Lawrence Berkeley Lab, NOT LANL. (See http://www.sciencenews.org/articles/20020720/fob5r ef.asp) And the mustang story is largely false, although the mainstream press did not make a big deal out of the fact that the story originally reported turned out to be untrue. (See http://www.lanl.gov/news/index.php?fuseaction=home .story&story_id=1453). Also keep in mind that that LBL, Argonne, Brookhaven, etc... do minimal amounts of classified work compared to LLNL, LANL, and SNL. Even PNL and ORNL do significantly less than the big three. So if LANL, LLNL, and SNL tend to have more security incidents, then one cannot ignore that a stellar record from a laboratory that does no classified work means very little in comparison.
Please get your facts right. It's that sort of uninformed, incorrect rhetoric and accusation that got LANL in the press-generated hot water it currently finds itself in. Are you a politician?
Odd -- I've seen the opposite when I've meta-moderated in the last few months. That, and people who appear to have very odd senses of humor judging by some of the +1 Funny mods out there...
I like iTunes also. It has a few features missing that I'd like, but I doubt they are interested in adding things that cater to folks with 100+GB music libraries. iTunes is good.
See, iPod != iTunes music store. It's a perfectly reasonable audio player that on it's own does NOT impose DRM rules on you (other than not being able to easily copy files OFF on someone elses computer). So it lacks OGG support -- yes, OGG sounds good (I don't know what the previous poster was smoking - OGG sounds fine). The iPod doesn't force you into anything even related to the iTunes music store if you don't want to use it.
And why would anyone want a "wealth of codecs"? I can't see why someone wouldn't just choose one, and stick with it, and possibly use a higher quality second choice for a subset of their music. My iPod supports a "wealth" of codecs already anyways - just not OGG.
"The only difference being that one group is employed by Microsoft, and the other likely do it in their spare time (they're usually still employed by some big-ass technology company)." Many of the big OSS projects (Linux kernel, Eclipse, as two huge examples) are predominantly written by people paid to work on the projects. The only difference between those hackers and one at microsoft is the openness of the code - not spare time vs work time. I think the "OSS = programmed in my spare time in the evening" argument went out the window a few years back for most of the truly large projects out there.
Using google hit counts is about as dumb as thinking that boxen is a real word simply by sticking a B on the front of the plural of Ox. "Common usage"? You do realize that the majority of computer users out there still refer to their computer as a "computer", not a "box" - let alone worrying about the plural of a somewhat niche-oriented, esoteric bit of terminology. To use your ill-informed google-metric to try to see how the "general population" uses the "word" "boxen", pages with references to boxen in the context of windows number in the mere 20,000 count. And I would speculate that many of those are uses by linux nerds who are in some way making nasty comments about Windows (disclaimer: linux user here. no windows junk in my space.)
Shall we remind you of the word "Jargon"? Existence in the world of jargon does not imply accepted use in the larger language.
Besides, FPGAs have two issues that make them good only for a very specific set of apps. Number 1, they don't currently have great floating point performance - this is a killer for most scientific apps. Number 2, they are hard to feed because the rate they can compute at versus the rate memory can feed them is quite skewed. Regardless, they're still very promising. The reconfigurable computing team at LANL (http://rcc.lanl.gov/) has done some very cool things with FPGA based systems.
Is anyone else frustrated when you hear wonderful science like this being done, yet see that probes like this are slated to have their funding cut (http://www.spacedaily.com/news/voyager1-05a.html) ? For some reason, $4.2 million / year to operate them (ie, listen) seems unbelievably cheap for such a unique resource - not only are there only TWO probes out there (voyager 1 and 2), but to get others out to replace them would cost a whole ton more. ...In addition to having to wait another 20 or so years to get there. :(
Science just doesn't work when politics gets involved...